Research Design and Measurement

Research Design and Measurement

1. Stages of Behavioural research

   • Problem
     • identifying your problem
   • Theory/hypothesis
     • Variables
       • independent (manipulated or predictor variable
       • Dependent (outcome) variable
   • Research design
     • Correlational/cross-sectional design
       • associations b/w 2 variables, no manipulation, can't establish causation
     • Longitudinal design
       • over time and multiple occasions in a group/cohort of people
     • Experimental design
       • measuring the effect of a factor on outcome of interest, can isolate variables
   • Measurement selection/design
   • Select sample
   • Collect data/run study
   • Code & analyse data
   • Write up findings

2. Validity and Reliability

   • Validity
     • the degree to which a measure accurately measure what it is supposed to measure
       • face validity - looks like it what it measures
       • Concurrent validity: a new measure that you may have developed correlate strongly with an already established measure
       • Predictive validity: does the measure actually measure the behaviour that it is supposed to represent (should predict results)
       • Discriminant validity: just measures what it is supposed to and not multiple factors
   • Reliability
     • Refers to the consistency or stability of a measure
       • Test-retest
         • Split-Half reliability

3. Forms of measurement

   • Nominal
     • labels, categories, or categorical date
   • Ordinal (scale)
     • ranking, ordering, relative position
   • Interval
     • assesses the distance between, no absolute zero
   • Ratio
     • Has an absolute zero, e.g. time passing from a starting point

4. Common problems in respondents

   • explicit (direct) measures may represent conscious (either true, or normative) attitudes
   • Implicit (indirect) attitudes may represent stored associations built upon over time - automatically activated

5. Common problems in measurement

   • Socially Desirable Responding (SDR)
     • impression management: we want to appear to be a good person, and normative influences (societal norms), mean that we answer in a way that doesn't truly represent our beliefs/attitudes/behaviours

6. Differences between explicit and implicit measures

   • Direct measures
     • self-report (explicit)
       • quantitative, scales, questionnaires, surveys, qualitative interviews)
   • Indirect measures
     • Observational
     • Physiological measures
       • GSR, EMG, EEG, fMRI
     • Implicit measures
       • word/stem completion, implicit associating test (IAT)